Emplois de quizá(s) et nouveaux contextes

This part of the analysis is based on flood maps developed using a combination of hydrological frequency analysis and hydrodynamic flood flow modelling, following a methodology previously applied³³ for national flood risk mapping in England. Firstly, the methods for statistical analysis of river flooding set out in the industry-standard

“Flood Estimation Handbook”³⁴ are applied to watercourses in the Thames catchment upstream of Kingston to derive estimates of flood flows at approximately every 200 metres along the stream network for five annual exceedance probabilities:

1/20, 1/75, 1/100, 1/200, 1/1000. The analysis includes all watercourses draining areas of more than 3 km². Then a hydrodynamic model is applied to simulate the limits of possible floodplain inundation (i.e. areas “at risk” of flooding) for each set of flow estimates. The software used, JFlow+, solves the two-dimensional depth-averaged shallow water equations with a finite volume implementation of Roe’s scheme^35,36 and has been demonstrated to be suitable for flood risk modelling in benchmark tests published by the official flood management authority in England³⁷. We apply it on a 5 m horizontal resolution grid with the ground elevations derived primarily from airborne LiDAR survey over the urban areas. The vertical resolution in LiDAR-derived terrain models is variable, but vertical root mean square errors are typically of the order of ~50mm³⁸.

Floodplain inundation is modelled for a notional world without flood defences, which would mitigate the actual risk in any specific flood event. This approximation, which we return to later, helps to assess the effects of climate forcing in isolation from other anthropogenic factors, and is consistent with the reporting of risk in official flood management plans³⁹. The resulting inundation maps are envelopes representing areas that could potentially be flooded with a given annual probability. Ordnance Survey “AddressPoint” data is then used to identify and count the properties within these areas. Supplementary Fig. 15 represents the number of properties thereby assessed to be at risk of flooding, with likelihood greater than the specified annual probability, in the absence of flood defences. By interpreting the annual exceedance probability of modelled river flows at Kingston as an index variable representing the severity of flooding in the catchment, Supplementary Fig. 15 is used as a lookup function to estimate, as a first approximation, how many properties could be at risk for any ensemble member.

Supplementary Figure 15: Number of properties individually at risk of flooding from the River Thames upstream of Kingston with annual probability greater than 1/T, not accounting for flood defences, as a function of return period T. Five scenarios were modelled (solid dots) for the specified river flow annual exceedance probabilities on watercourses draining sub-catchments larger than 3 km².

This relationship is adopted as an approximate impact function, applied so as to obtain an indication of the number of properties flooded in each of over 130,000 ensemble simulations of a complex hydro-meteorological model chain. It is acknowledged that this does not account for uncertainties in the flood inundation modelling process, nor the effect of biases in the outputs of the hydro-meteorological modelling chain relative to actual extreme flows in the Thames catchment. A comprehensive uncertainty analysis of the entire modelling chain would ideally be performed, but was not feasible in the present study. However the property counts for the Actual Conditions simulations are broadly in line with the Environment Agency’s Thames Catchment Flood Management Plan³⁴, which estimated that approximately 135,000 properties would have more than a 1-in-100 chance of riverine flooding in any one year, without flood defences. That figure differs in detail from the estimates adopted here because it is based on a composite of several inundation model outputs, and also different property datasets and property counting assumptions.

Return time [years]

Number of properties at risk (x 103 )

1 10 100 1000

050100150200

Modelled scenarios Assumed zero Interpolated Extrapolated

To assess the difference in the number of properties at risk of flooding between Actual Conditions and Natural, the frequency distributions of the simulated river flows at Kingston are derived from the hydrological model outputs for the Actual Conditions case, and for each of the Natural ensembles. For each ensemble, the Natural forcing river flows Q expressed on the physical river flow scale, are compared with the distribution of flows from the Actual Conditions simulations, GA(Q), to calculate the corresponding annual probabilities of exceedance 1 - GA(Q) on the Actual Conditions scale. This effectively translates the empirical distribution of peak flows from the Natural ensembles onto the same scale as the Actual Conditions simulations, allowing the relationship shown in Supplementary Fig.15 to be used to estimate the change in number of properties at risk for return times on the Actual Conditions scale, as shown in Fig. 5f.

Flood protection measures within the Thames river basin have evolved as a complex mixture of raised embankments, artificially straightened drainage channels, river diversions and other structures. Official flood management plans⁴⁰ describe how the geology of the Thames floodplain makes construction of raised flood defences impractical in many places, and show that although there are numerous assets acting to reduce flood risk, only 3%⁴⁰ of the total floodplain area is classified as being protected by “significant” flood defences, benefitting 5% of properties that would otherwise be at risk of flooding with a 1% or greater annual probability. Some 10% of the floodplain is classed as heavily populated and not protected by flood defences, and these areas contain around 40% of properties at risk (numbering 56,000).

Approximately 69% of the Thames floodplain (or 14% of properties at risk) is classed as being in “open floodplain”, which includes a mixture of defended and undefended areas. Neglecting the role of flood defences is thus considered a reasonable approximation for the purposes of this analysis.

Sensitivity of the estimated change in risk to the assumptions made about flood defences can be assessed in terms of the average annual economic cost of flooding.

The annual average flood damage for a typical UK residential property without protection is estimated⁴¹ to be £4,947 (at 2015/16 prices), hence the annual economic cost associated with the changes in risk attributable to human-induced climate change in this study can be estimated as between approximately -£19.8 million (a reduction corresponding to 4000 fewer properties at risk) and +£39.6

million (an increase corresponding to 8,000 more properties at risk). The most favourable standard of protection for areas benefitting from “significant” defences in the Thames catchment is reported to be 1/200 annual probability⁴⁰, for which the average annual damages of a typical property reduce⁴¹ to £40. Assuming that flood defences of this standard would have benefitted the same proportion of properties in any of the Natural ensembles as in the actual catchment (i.e. 5% of properties, see above), then the upper bound of the change in risk attributable to climate change would be reduced by £1.96 million to £37.6 million, a relatively insignificant reduction.

The results presented here are intended as a realistic indication of the potential flood risk, under different climatic forcing scenarios, based on detailed contemporary flood mapping and property data. Inputs to CLASSIC are spatially distributed on a grid, as are its internal runoff calculations, but the runoff is then routed to the catchment outlet at Kingston in order to predict the river flow there, which is the primary model output. In the absence of spatially distributed estimates of river flow, the return period T (years) of the daily peak river flows at Kingston is applied as an indicator of the relative extremeness of flooding throughout the catchment. This approximation neglects the spatio-temporal details of individual events, but is consistent with the strong spatial dependence in extreme river flows in this catchment, especially for prolonged flood events in the winter season⁴².

Also the figures are based on a recent snapshot of properties in the Thames region, which is assumed to be a fixed representation of the built environment. The analysis therefore takes no account of how property development might have differed under climate conditions consistent with the Natural forcing.

The results are based on statistical analysis of peak river flows and a robust, physics-based floodplain model applied at a relatively high spatial resolution.

However, the modelling necessarily involves some approximation of the real flood risk in the Thames catchment. A further, more comprehensive analysis of potential flood damage for the Thames region might be able to take into account additional factors, including:

• The specific locations, standards and performance of flood defence systems

• Variation in the spatial extent and timing of flood events

• The evolution and duration of flooding within an event

• The risk associated with sea surge in the tidal Thames (i.e. “downstream” of Kingston)

• Surface water flooding associated with overland runoff and the performance of surface and sub-surface drainage systems

• Groundwater levels

At present the integration of these factors in assessments of flood risk remains a challenge both for researchers and for the flood risk management industry.

References

1 Kistler, R. et al. The NCEP-NCAR 50-year reanalysis: Monthly means CD-ROM and documentation. Bull. Amer. Meteorol. Soc. 82, 247-267,

doi:10.1175/1520-0477(2001)082<0247:tnnyrm>2.3.co;2 (2001).

2 Compo, G. P. et al. The Twentieth Century Reanalysis Project. Quarterly Journal of the Royal Meteorological Society 137, 1-28, doi:10.1002/qj.776 (2011).

3 Trenberth, K. E. & Paolino, D. A. Characteristic Patterns of Variability of Sea Level Pressure in the Northern Hemisphere. Mon. Weather Rev. 109, 1169-1189,

doi:10.1175/1520-0493(1981)109<1169:CPOVOS>2.0.CO;2 (1981).

4 Massey, N. et al. weather@home - development and validation of a very large ensemble modelling system for probabilistic event attribution. Quarterly Journal Of The Royal Meteorological Society In press, doi:10.1002/qj.2455 (2014).

5 Stark, J. D., Donlon, C. J., Martin, M. J. & McCulloch, M. E. OSTIA : An operational, high resolution, real time, global sea surface temperature analysis system. Oceans 2007 - Europe, Vols 1-3, 331-334 (2007).

6 Donlon, C. J. et al. The Operational Sea Surface Temperature and Sea Ice Analysis (OSTIA) system. Remote Sens. Environ. 116, 140-158 (2012).

7 Schaller, N. et al. The heavy precipitation event of May-June 2013 in the upper Danube and Elbe basins. Bull. Amer. Meteorol. Soc. 95, S69-S72 (2014).

8 Taylor, K. E., Stouffer, R. J. & Meehl, G. A. An Overview of CMIP5 and the Experiment Design. Bull. Amer. Meteorol. Soc. 93, 485-498 (2012).

9 Dee, D. P. et al. The ERA-Interim reanalysis: configuration and performance of the data assimilation system. Quarterly Journal of the Royal Meteorological Society 137, 553-597, doi:10.1002/qj.828 (2011).

10 von Storch, H. & Zwiers, F. W. Statistical Analysis in Climate Research. (Cambridge University Press, 2001).

11 Michelangeli, P. A., Vautard, R. & Legras, B. Weather regimes - Reccurence and quasi stationarity. J. Atmos. Sci. 52, 1237-1256,

doi:10.1175/1520-0469(1995)052<1237:wrraqs>2.0.co;2 (1995).

12 Peixoto, J. P. & Oort, A. H. Physics of Climate. (American Institute of Physics, 1992).

13 Crooks, S. M. & Naden, P. S. CLASSIC: a semi-distributed rainfall-runoff modelling system.

Hydrol. Earth Syst. Sci. 11, 516-531 (2007).

14 Bell, V. A. & Moore, R. J. An elevation-dependent snowmelt model for upland Britain.

Hydrological Processes 13, 1887-1903, doi:10.1002/(SICI)1099-1085(199909)13:12/13<1887::AID-HYP846>3.0.CO;2-C (1999).

15 Prudhomme, C., Crooks, S., Kay, A. L. & Reynard, N. Climate change and river flooding: part 1 classifying the sensitivity of British catchments. Climatic Change 119, 933-948,

doi:10.1007/s10584-013-0748-x (2013).

16 Prudhomme, C. et al. Future Flows Hydrology: an ensemble of daily river flow and monthly groundwater levels for use for climate change impact assessment across Great Britain. Earth Syst. Sci. Data 5, 101-107, doi:10.5194/essd-5-101-2013 (2013).

17 Kay, A. L., Crooks, S. M., Pall, P. & Stone, D. A. Attribution of Autumn/Winter 2000 flood risk in England to anthropogenic climate change: A catchment-based study. Journal of Hydrology 406, 97-112, doi:10.1016/j.jhydrol.2011.06.006 (2011).

18 Morris, D.G. and Flavin, R.W. (1990). A digital terrain model for hydrology. Proceedings of the 4th International Symposium on Spatial Data Handling, Zurich, Volume 1, 250-262.

19 Marsh, T. & Harvey, C.L. 2012. The Thames flood series: a lack of trend in flood magnitude and a decline in maximum levels. Hydrology Research, 43(3), 203-214

20 Bell, V. A. et al. How might climate change affect river flows across the Thames Basin? An area-wide analysis using the UKCP09 Regional Climate Model ensemble. Journal of Hydrology 442, 89-104, doi:10.1016/j.jhydrol.2012.04.001 (2012).

21 Huntingford, C. et al. Potential influences on the United Kingdom's floods of winter 2013/14.

Nature Climate Change 4, 769-777, doi:10.1038/nclimate2314 (2014).

22 Keller, V. D. J. et al. CEH-GEAR: 1 km resolution daily and monthly areal rainfall estimates for the UK for hydrological use. Earth Syst. Sci. Data 7, 143-155, doi:10.5194/essd-7-143-2015 (2015).

23 Hough, M. N. & Jones, R. J. A. The United Kingdom Meteorological Office rainfall and

evaporation calculation system: MORECS version 2.0-an overview. Hydrol. Earth Syst. Sci. 1, 227-239 (1997).

24 Jenkins, G. J., Perry, M. C. & Prior, M. J. The climate of the United Kingdom and recent trends. (Met Office Hadley Centre, Exeter, UK, 2007).

25 Crooks, S.M. & Kay, A.L. Simulation of river flow in the Thames over 120 years: Evidence of change in rainfall-runoff response? Journal of Hydrology: Regional Studies, 4, 172-195, doi:10.1016/j.ejrh.2015.05.014 (2015).

26 Oudin, L. et al. Which potential evapotranspiration input for a lumped rainfall-runoff model?

Part 2 - Towards a simple and efficient potential evapotranspiration model for rainfall-runoff modelling. Journal of Hydrology 303, 290-306, doi:10.1016/j.jhydrol.2004.08.026 (2005).

27 Kay, A.L. & Davies, H.N. (2008). Calculating potential evaporation from climate model data: a source of uncertainty for hydrological climate change impacts. Journal of Hydrology 358, 221–

239, doi: 10.1016/j.jhydrol.2008.06.005.

28 Prudhomme, C. & Williamson, J. (2013). Derivation of RCM-driven potential

evapotranspiration for hydrological climate change impact analysis in Great Britain: a

comparison of methods and associated uncertainty in future projections. Hydrology and Earth System Sciences, 17, 1365–1377, doi:10.5194/hess-17-1365-2013.

29 Kay, A.L., Bell, V.A., Blyth, E.M., Crooks, S.M., Davies, H.N. and Reynard, N.S. A

hydrological perspective on evaporation: historical trends and future projections in Britain.

Journal of Water and Climate Change, 4(3), 193–208, doi:10.2166/wcc.2013.014 (2013).

30 Kay, A. L., Reynard, N. S. & Jones, R. G. RCM rainfall for UK flood frequency estimation. I.

Method and validation. Journal of Hydrology 318, 151-162, doi:10.1016/j.jhydrol.2005.06.012 (2006).

31 Gringorten, I. I. A plotting rule for extreme probability paper. Journal of Geophysical Research 68, 813-814, doi:10.1029/JZ068i003p00813 (1963).

32 Allen, M. Liability for climate change. Nature 421, 891-892 (2003).

33 Bradbrook, K., Waller, S., & Morris, D. National floodplain mapping: Datasets and methods - 160,000 km in 12 months. Natural Hazards, 36(1-2), 103-123 (2005).

34 Flood Estimation Handbook. (Centre for Ecology and Hydrology, 1999).

https://www.gov.uk/government/publications/thames-catchment-flood-management-plan 35 Roe, P.L. and Pike, J. Efficient construction and utilisation of approximate Riemann solutions.

In: Glowinski, R. and Lions, J.L. (Eds). Proc. of the Sixth Int. Symposium on Computer Methods in Applied Sciences and Engineering, 499-518, Amsterdam: North-Holland (1984).

36 Toro, E.F. Shock–Capturing Methods for Free-Surface Shallow Flows. Wiley 326 pp. (2001).

37 Benchmarking the latest generation of 2D hydraulic flood modelling packages. Environment Agency Report SC120002, ISBN: 978-1-84911-306-9, Bristol, UK (2013).

http://evidence.environment-agency.gov.uk/FCERM/Libraries/FCERM_Project_Documents/SC120002_Benchmarking_2D _hydraulic_models_Report.sflb.ashx (Accessed August 2015)

38 Fewtrell, Timothy J., Duncan, Alastair, Sampson, Christopher C., Neal, Jeffrey C., & Bates, Paul D. (2011). Benchmarking urban flood models of varying complexity and scale using high resolution terrestrial LiDAR data.Physics and Chemistry of the Earth, 36(7), 281-291.

doi:10.1016/j.pce.2010.12.011

39 Thames Catchment Flood Management Plan. Environment Agency ref: LIT 4403, GETH1209BQYL-E-E, Bristol, UK (2009).

40 Thames Catchment Flood Management Plan. Environment Agency ref: LIT 4403, GETH1209BQYL-E-E, Bristol, UK (2009).

41 Penning-Rowsell, E., Priest, S., Parker, D., Morris, J., Tunstall, S., Viavattene, C., Chatterton, J. and Owen, D. Handbook for Economic Appraisal: Flood and Coastal Erosion Risk Management (Flood Hazard Research Centre, 2015).

42 Keef, C., Svensson, C. and Tawn, J. A. Spatial dependence in extreme river flows and precipitation for Great Britain. J. Hydrology 378, 240-252, doi: 10.1016/j.jhydrol.

2009.09.026 (2009).